CK2 activity is required for the interaction of FGF14 with voltage-gated sodium channels and neuronal excitability

Abstract

Recent data shows that fibroblast growth factor 14 (FGF14) binds to and controls the function of the voltage-gated sodium (Nav) channel with phenotypic outcomes on neuronal excitability. Mutations in the FGF14 gene in humans have been associated with brain disorders that are partially recapitulated in Fgf14^−/− mice. Thus, signaling pathways that modulate the FGF14:Nav channel interaction may be important therapeutic targets. Bioluminescence-based screening of small molecule modulators of the FGF14:Nav1.6 complex identified 4,5,6,7-tetrabromobenzotriazole (TBB), a potent casein kinase 2 (CK2) inhibitor, as a strong suppressor of FGF14:Nav1.6 interaction. Inhibition of CK2 through TBB reduces the interaction of FGF14 with Nav1.6 and Nav1.2 channels. Mass spectrometry confirmed direct phosphorylation of FGF14 by CK2 at S228 and S230, and mutation to alanine at these sites modified FGF14 modulation of Nav1.6-mediated currents. In 1 d in vitro hippocampal neurons, TBB induced a reduction in FGF14 expression, a decrease in transient Na⁺ current amplitude, and a hyperpolarizing shift in the voltage dependence of Nav channel steady-state inactivation. In mature neurons, TBB reduces the axodendritic polarity of FGF14. In cornu ammonis area 1 hippocampal slices from wild-type mice, TBB impairs neuronal excitability by increasing action potential threshold and lowering firing frequency. Importantly, these changes in excitability are recapitulated in Fgf14^−/− mice, and deletion of Fgf14 occludes TBB-dependent phenotypes observed in wild-type mice. These results suggest that a CK2-FGF14 axis may regulate Nav channels and neuronal excitability.—Hsu, W.-C. J., Scala, F., Nenov, M. N., Wildburger, N. C., Elferink, H., Singh, A. K., Chesson, C. B., Buzhdygan, T., Sohail, M., Shavkunov, A. S., Panova, N. I., Nilsson, C. L., Rudra, J. S., Lichti, C. F., Laezza, F. CK2 activity is required for the interaction of FGF14 with voltage-gated sodium channels and neuronal excitability.

The intracellular fibroblast growth factors (iFGFs) FGF11–14 are a family of multivalent accessory proteins that interact with the macromolecular voltage-gated sodium channel (Nav) 1.1–1.9 complex, a key molecular determinant of neuronal excitability (1, 2). Disruption of iFGF binding to Nav channels has been shown to adversely affect firing in cardiac and neuronal cells (3–6). In particular, FGF14, an emerging potential risk gene for schizophrenia, depression, and anxiety (7) and the recognized genetic cause of hereditary spinocerebellar ataxia (8), strongly binds to the intracellular C-terminus of the α subunits of Nav1.1, 1.2, and 1.6 (9), the 3 dominant Nav channel isoforms in the brain. In rat hippocampal neurons, heterologous FGF14 colocalizes with native Nav at the axonal initial segment (AIS) (10), a specialized domain containing a high density of Nav channels responsible for action potential (AP) initiation and propagation (11). Disruption of FGF14 reduces Nav α subunit expression at the AIS, reduces Na⁺ channel current density, and reduces excitability of hippocampal neurons (5, 12). In animal studies, fgf14^−/− mice exhibit decreased excitability (13, 14), impaired synaptic transmission (15), and consequently deficits in motor and cognitive skills (16, 17). Additionally, FGF14 controls cellular targeting of Nav channels located at the AIS (4, 12), the subcellular domain required for AP initiation (18, 19), through a glycogen synthase kinase 3 (GSK-3)-centered network (12, 20, 21), which may regulate targeting of this complex to the AIS and modulate neuronal excitability (12, 20). The multimodal regulation of the FGF14:Nav complex by protein kinases has spurred great interest in investigating FGF14-convergent signaling pathways that may produce rapid, fine-grained regulation of excitability.

A critical member of the intracellular signaling kinome in neurons is casein kinase 2 (CK2), a serine/threonine protein kinase comprised of a tetramer with 2 catalytic α and 2 regulatory β subunits. The catalytic domains of CK2α and CK2β are highly conserved, suggesting tight regulation of in vivo CK2 activity (22). One of the primary functions of CK2 in neurons is to serve as a priming kinase for GSK-3. Phosphorylation of the S/T site downstream of the GSK-3 motif (S/TXXS/T; S/TXXXS/T, the CK2 priming site is shown in bold) (23) enhances GSK-3 phosphorylation and amplifies the GSK-3 signaling cascade (24). Additionally, CK2 phosphorylates Nav channels at S1112, S1124, and S1126, residues that are within the Ankyrin-G binding site (25, 26); phosphorylation at these sites regulates trafficking to and stability of Nav channels at the AIS (27). Thus, prolonged inhibition of CK2 activity by pharmacological inhibitors (>24 h), genetic silencing, or overexpression of dominant-negative Nav channel phosphosilent constructs disrupts Nav channels localization at the AIS and impairs formation of neuronal polarity in immature neurons through these sites (28–30). In addition to Nav channels, CK2 also phosphorylates the schwannomin-interacting protein IQCJ-SCHIP-1, which mediates the association of the ankyrin-binding motif of Nav channels with Ankyrin-G (31), a crucial intracellular protein for the maintenance of the AIS and nodes of Ranvier (32). Likewise, Nav channel expression mediates CK2 clustering at the AIS (25, 26). However, the role of CK2 in the regulation of iFGFs has not been previously shown.

Based on convergent evidence for CK2 as a member of the AIS and a regulator of GSK-3 activity, we postulated the existence of signaling crosstalk between CK2 and FGF14. To this end, we built on previous high-throughput screening studies of kinase inhibitors (9, 12, 20, 33, 34), tested for their ability to modulate the FGF14:Nav1.6 complex, and identified a subset of compounds that target CK2. Through a combination of luminescence-based validation assays, followed by orthogonal validation through coimmunoprecipitation, in vitro phosphorylation followed by mass spectrometry, confocal microscopy, and patch-clamp electrophysiology, we demonstrate a novel role of CK2 in controlling FGF14 assembly to Nav channels, and in phosphorylating FGF14. TBB (4,5,6,7-tetrabromobenzotriazole), a potent CK2 inhibitor, rapidly abolishes the FGF14:Nav1.6 interaction and reduces the ability of FGF14 to bind to Nav1.6 and Nav1.2. CK2 phosphorylates FGF14 at S228 and S230 in vitro, and TBB reduces FGF14 expression at days in vitro (DIV) 1 neurons and causes a time-dependent redistribution of FGF14 from the AIS in mature neurons. Furthermore, TBB suppresses the amplitude of Na⁺ currents and induces a hyperpolarizing shift in voltage dependence of steady-state inactivation of Nav channels in hippocampal neurons. In brain slices, CK2 inhibition disrupts intrinsic excitability of cornu ammonis area 1 (CA1) hippocampal pyramidal neurons by increasing AP current threshold and impairing neuronal excitability, a phenotype occluded by genetic deletion of Fgf14. These results provide evidence for a new mechanism of regulation of the macromolecular complex of the Nav channel and neuronal excitability through a functionally relevant CK2-FGF14 axis.

MATERIALS AND METHODS

DNA constructs

The Cluc-FGF14, cluster of differentiation 4 (CD4)-Nav1.6-Nluc, and FGF14-6×myc plasmids used in this study were previously described (9, 12). The FGF14^S228A/S230A-GFP construct was generated by using the QuikChange Lightning kits protocol (Agilent Technologies, Santa Clara, CA, USA) using the primers (5′→3′ orientation) GACGCCAAGTAAAGCCACAAGTGCGTCT (forward) and CTGCGGTTCATTTCGGTGTTCACGCAGA (reverse) for S228A, and CAAGTAAAGCCACAGCTGCGTCTGCAAT (forward) and GTTCATTTCGGTGTCGACGCAGACGTTA (reverse) for S230A. Constructs were confirmed by sequencing at Molecular Genomics Core Facility [University of Texas Medical Branch (UTMB), Galveston, TX, USA].

Chemicals

Chemicals and antibodies used are detailed in Supplemental Table S1.

Cell culture and transient transfections

Human embryonic kidney 293 (HEK293) cells were maintained in DMEM (Invitrogen, Carlsbad, CA, USA), supplemented with 10% FBS, 100 U/ml penicillin, 100 µg/ml streptomycin, and incubated at 37°C with 5% CO2. To HEK293 cells stably expressing rat Nav1.2 (HEK-Nav1.2) and rat Nav1.6 (HEK-Nav1.6), 500 µg/ml G418 (Invitrogen) were added to maintain stable transfectants. For transient transfections, HEK293 cells were transfected at 90–100% confluency using Lipofectamine 2000 (Invitrogen), according to manufacturer’s instructions.

Bioluminescence assays

Assays were performed as previously described (9, 12, 20, 33, 34). Briefly, HEK293 cells (∼4.5 × 10⁵ per 15.5 mm diameter well in 24-well plates) were transiently cotransfected with pairs of plasmids or single plasmids using Lipofectamine 2000 (Invitrogen), transferred to 96-well white-walled plates 24 h prior to luminescence readings, and incubated 30 min in serum-free, phenol-red free DMEM/F12 prior to treatment with growth medium containing appropriate drugs or vehicles for the time indicated. Luminescence readings were performed with a Synergy H4 Reader (BioTek, Winooski, VT, USA) initiated by automated injection of 100 μl of substrate (in a 1:1 volume ratio) containing 1.5 mg/ml of d-luciferin (final concentration of 0.75 mg/ml), followed by 3 s of mild plate shaking, and measurements taken at 2 min intervals with 0.5 s integration time for a total duration of 30 min. Raw signal intensity was computed by averaging peak luminescence plus 2 adjacent time points. Normalized signal intensity was expressed as percentage of mean signal intensity relative to control treated with 0.5% DMSO. Full-length Photinus luciferase activity was determined by transfecting HEK293 cells, as above, with pGL3 firefly luciferase plasmid, as described previously (34). Cell viability was assessed using a CellTiter Cell Proliferation Assay (MTT assay) kit in the presence of 0.5% DMSO or inhibitor according to manufacturer instructions (Promega, Madison, WI, USA).

Peptide synthesis

The peptide FGF14 [KPGVTPSKSTSASAIMNGGK-NH₂] (MW = 1917.20 Da) was synthesized on a CEM Liberty Blue Discover automated microwave synthesizer (Matthews, NC, USA). Solid phase coupling was completed on Rink-amide resin using standard hydroxybenzotriazole/diisopropylcarbodiimide chemistries. Peptides were cleaved and lyophilized as previously described (21). Peptides were purified by reverse phase HPLC over a gradient of 20–60% acetonitrile:H₂O over 60 min. Final product molecular weights were verified by MALDI-MS and purity was determined to be >95% by HPLC.

In vitro phosphorylation and sample preparation

CK2 in vitro phosphorylation of the FGF14 peptide [KPGVTPSKSTSASAIMNGGK-NH₂] using baculovirus-produced CK2α1 (Promega) was performed according to manufacturer’s instructions. Control studies were performed under identical conditions, but lacking the addition of the kinase to the reaction solution. Samples were reduced, alkylated, digested, desalted, and lyophilized as previously described (21).

Mass spectrometry and data analysis

Samples were analyzed by nanoLC-MS/MS using an Orbitrap Elite mass spectrometer (Thermo Scientific, Rockford, IL, USA) as previously described (21). Sample loading, solvent gradients, and acquisition parameters were unchanged. Mass spectrometry (.raw) data files were imported into the Peaks software package (version 6; Bioinformatics Solutions Inc., Waterloo, ON, Canada) to search and identify phosphorylation sites. Parent ion tolerance was set to 10 ppm, and fragment tolerance was set to 0.1 Da with a maximum 2-miss cleavage allowance. Identified phosphopeptide spectra were manually annotated using the MS-Product tool on the Protein Prospector website (prospector.ucsf.edu) to generate theoretical m/z values for all fragment ions. Phosphorylation sites were identified manually by locating all present site-identifying b and y ions in the sequence.

Western blotting

HEK-Nav1.2 or HEK-Nav1.6 cells were treated for 1 h at 37°C with kinase inhibitors (or DMSO), then washed with PBS and buffer containing (in mM): 20 Tris-HCl, 150 NaCl, 1% NP-40. Lysis was performed by 20 s sonication with addition of protease inhibitor cocktail (set 3, Calbiochem), and centrifuged at 4°C, 15,000 g for 15 min, adding 4× sample buffer containing 50 mM tris-(2-carboxyethyl)phosphine (TCEP) (Sigma-Aldrich, St. Louis, MO, USA). Mixtures were heated for 10 min at 65°C, resolved on 4–15% polyacrylamide gels (Bio-Rad, Hercules, CA, USA), and transferred to PVDF membranes as previously described (12). For Western blot, membranes were then incubated in blocking buffer (TBS with 3% nonfat dry milk and 0.1% Tween-20) containing mouse monoclonal anti-Luciferase mAb (1:1000, Sigma-Aldrich), mouse monoclonal anti-myc (1:1000; 9E10 clone Santa Cruz Biotechnology, Santa Cruz, CA, USA), or mouse monoclonal anti-PanNav channel (1:1000; Sigma Aldrich) antibody overnight. Washed membranes were incubated with goat anti-mouse or goat anti-rabbit HRP antibody (1:4000–8000; Thermo Scientific) and detected with ECL Advance Western Blotting Detection kit (GE Healthcare, Piscataway, NJ, USA); protein bands were visualized using FluorChem HD2 System and analyzed with AlphaView 3.1 software (ProteinSimple, Santa Clara, CA, USA).

Immunoprecipitations

Immunoprecipitations from HEK-Nav1.2 and HEK-Nav1.6 cells were performed as previously described (12). Cells were washed twice with PBS and lysed in the following lysis buffer: 20 mM Tris-HCl, 150 mM NaCl, and 1% NP-40 or Triton X-100. Protease inhibitor mixture (set 3, Calbiochem) was added immediately before cell lysis. Cell extracts were collected and sonicated for 20 s and centrifuged at 4°C, at 15,000 g for 15 min. Supernatants were collected and incubated with rabbit anti-myc agarose beads (Sigma-Aldrich) for 2 h at 4°C with agitation. After washing 5 times with lysis buffer, 2× sample buffer (Bio-Rad) containing 50 mM TCEP was added. Lysates were then heated for 10–15 min at 65°C and resolved on 7.5 or 4–15% polyacrylamide gradient gels (Bio-Rad). Resolved proteins were transferred to PVDF (Millipore, Bedford, MA, USA) for 2 h at 4°C and blocked in Tris-buffered saline with 5% skim milk and 0.1% Tween 20. Membranes were then incubated in blocking buffer containing a monoclonal anti-myc (1:1000; Santa Cruz Biotechnology) or anti-PanNav channel (1:1000; Sigma-Aldrich) antibody overnight at 4°C. Washed membranes were incubated with goat anti-mouse HRP (1:5000–10,000) detected with ECL Advance Western Blotting Detection kit (GE Healthcare). Protein bands were visualized using FluorChem HD2 System and analyzed with AlphaView 3.1 software (ProteinSimple).

Primary neuronal cultures and immunocytochemistry

Banker’s style hippocampal neuron cultures were prepared from embryonic d 18 rat embryos using previously described methods (9, 12). Hippocampal neurons (DIV1–2 or DIV21) were treated with TBB (30 μM for DIV1–2; 50 μM for DIV21) or DMSO (0.25%) for various times (2 h for DIV1–2; 8–12 h for DIV21), then fixed in 4% paraformaldehyde/sucrose for 15 min. After permeabilization with 0.25% Triton X-100 for 5 min, coverslips were washed with PBS, blocked in 10% bovine serum albumin/PBS for 1 h at 37°C, and incubated with combinations of the following primary antibodies: mouse anti-FGF14 IgG1 (monoclonal, 1:100; University of California, Davis/National Institutes of Health (NIH) NeuroMab Facility, Davis, CA, USA), mouse anti-PanNav IgG1 (monoclonal, 1:100; Sigma-Aldrich), chicken anti-βIV-spectrin IgG (polyclonal, 1:5000; gift from Dr. Komada, Tokyo Institute of Technology, Tokyo, Japan), chicken anti-MAP2 (polyclonal, 1:5000; Covance, NJ, USA). Samples were then washed, incubated with combinations of isotype-specific Alexa 488, 568, and 647 (1:200) secondary antibodies for 1 h at 37°C, washed again, and mounted onto glass microscope slides with Prolong Gold anti-fade reagent (Invitrogen).

Image analysis

Confocal images were acquired with a Zeiss LSM-510 confocal microscope with a ×63 oil-immersion objective (1.4 numerical aperture; Carl Zeiss GmbH, Jena, Germany) as previously described (12). Multitrack acquisition was performed with excitation lines at 488, 543, and 633 nm for Alexa 488, Alexa 568, and Alexa 633, respectively. Respective emission filters were bandpass 505–530 nm, 560–615 nm, and low-pass 650 nm. The optical slices were 0.8 μm; z stacks were collected at z steps of 0.4 μm with a frame size of 1024 × 1024, pixel time of 2.5 μs, pixel size 0.1×0.1 μm (×63) and an 8 frame Kallman averaging. Acquired z stacks were sum-projected and analyzed using ImageJ 1.48 s (NIH, Bethesda, MD, USA). Regions of interest were drawn as previously described (12).

Patch-clamp electrophysiology in HEK293 cells

HEK293 cells stably expressing Nav1.6 (HEK-Nav1.6) were transfected with FGF14-GFP or FGF14^S228A/S230A-GFP, dissociated, and replated at low density ∼12 h post-transfection. Recordings were performed at room temperature (20–22°C) 12–18 h post-transfection using a Multiclamp 700B Amplifier (Molecular Devices, Sunnyvale, CA, USA). Borosilicate glass pipettes with resistance of 3–6 MΩ were made Narishige International Inc. (East Meadow, NY, USA). The recording solutions were as follows: extracellular (mM): 140 NaCl, 3 KCl, 1 MgCl2, 1 CaCl2, 10 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 10 glucose, pH 7.3; intracellular: 130 CH3O3SCs, 1 EGTA, 10 NaCl, 10 HEPES, pH 7.3. Membrane capacitance and series resistance were estimated by the dial settings on the Multiclamp 700B amplifier (Molecular Devices). Capacitive transients and series resistance were compensated electronically by 75–85%. Cells with series resistance more than 25 mΩ were not considered for analysis. Data were acquired at 20 kHz and filtered at 2.2 kHz prior to digitization and storage. All experimental parameters were controlled by Clampex 9 software (Molecular Devices) and interfaced to the electrophysiological equipment using a Digidata 1322A analog-digital interface (Molecular Devices). Voltage-dependent inward currents for HEK-Nav1.6 cells were evoked by 125 ms depolarizing pulses with 5 s interpulse-interval to test potentials between −100 and +60 mV (5 mV increments) from a holding potential of −70 mV followed by 500 ms voltage prestep pulse of −120 mV. Steady-state inactivation of Nav channels was measured with a paired-pulse protocol. From holding potential of −70 mV, cells were given 645 ms voltage step of varying potentials between −120 and +20 mV (prepulse) prior to 50 ms test pulse to −20 mV with 5 s interpulse interval.

Patch-clamp electrophysiology in cultured neurons

Whole-cell patch-clamp recordings were obtained from cultured rat hippocampal neurons at DIV1–3 at room temperature (20–22°C) using a MultiClamp 700B amplifier (Molecular Devices), low-pass filtered at 2.2 kHz, and sampled at 20 kHz using a Digidata 1322A analog-to-digital interface and pClamp9 acquisition software (Molecular Devices). Borosilicate glass pipettes with resistance of 4–5 MΩ were made using a Narishige PC10 vertical Micropipette Puller (Narishige International Inc.). The extracellular bath solution contained (in mM) 140 NaCl, 3 KCl, 1 MgCl₂, 1 CaCl₂, HEPES, and 10 glucose, pH 7.4; the intracellular solution contained the following (in mM): 130 CH₃CsO₃S, 1 EGTA, 10 NaCl, 10 HEPES, pH 7.3, adjusted with CsOH. Bicuculline (20 µM), 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo[f]quinoxaline-2,3-dione (NBQX) (20 µM), and (2R)-amino-5-phosphonovaleric acid (APV) (100 µM) were added to block synaptic activity mediated via GABA and glutamate receptors subtypes AMPA and NMDA, respectively. Capacitive transients and series resistance were compensated electronically by 75–85%. Cells with series resistance more than 25 mΩ were not considered for analysis. Cells with series resistance more than 25 mΩ were not considered for analysis. Voltage-dependent inward sodium currents were evoked by 125 ms depolarizing voltage step with 5 s interpulse-interval to test potentials between −60 and +70 mV (10 mV increment) from a holding potential of −70 mV followed by a 500 ms voltage prestep pulse of −90 mV. Steady-state inactivation of Nav channels was measured with a paired-pulse protocol. From the holding potential, cells were stepped to 645 ms prepulse potentials between −110 and +20 mV and interpulse-interval of 5 s prior to 50 ms test pulse of −10 mV. Current densities, current-voltage, steady-state activation, and steady-state inactivation curves were derived as described previously (12).

Animal colonies

Fgf14^−/− and Fgf14^+/+ male and female mice were maintained on an inbred C57/BL6J background and bred in the UTMB animal care facility using heterozygous Fgf14^+/− males and females; Fgf14^+/+ wild-type mice served as control. Both male and female mice were used in this study at 2–3 mo of age, unless otherwise stated. The University of Texas Medical Branch operates in compliance with the U.S. Department of Agriculture Animal Welfare Act, the NIH Guide for the Care and Use of Laboratory Animals, and American Association for Laboratory Animal Science, Institutional Animal Care and Use Committee-approved protocols. Mice were housed, n ≤ 5 per cage, kept under a 12 h light–12 h dark cycle with sterile food and water ad libitum. All genotypes described were confirmed by genotyping of the progeny using DNA extraction and PCR amplification was conducted at Transnetyx, Inc. (Cordova, TN, USA).

Hippocampal slice preparation and whole-cell patch clamp

Hippocampal coronal slices (300–400 μm thick) from age-matched mice (2–3 mo, n = 4 per group, n = 18–21 independent experiments) were prepared using standard procedures (35). After decapitation the brains were rapidly removed and placed in ice-cold cutting solution containing in mM: 124 NaCl, 3.2 KCl, 1 NaH₂PO₄, 26 NaHCO₃, 2 MgCl₂, 1 CaCl₂, 10 glucose, 2 Na-pyruvate, and 0.6 ascorbic acid (pH 7.4, 95%O₂/5% CO₂). Slices were cut with a vibratome (VT1200 S; Leica Microsystems, Wetzlar, Germany) and incubated in the cutting solution at 32°C for at least 1 h and then at room temperature until use. Slices were then transferred to a submerged recording chamber and continuously perfused with artificial cerebrospinal fluid bubbled with 95% O2 and 5% CO2 (pH 7.4) containing (in mM): 124 NaCl, 3.2 KCl, 1 NaH₂PO₄, 26 NaHCO₃, 1 MgCl₂, 2 CaCl₂, 10 glucose. After at least 60 min of recovery, recordings were performed in artificial cerebrospinal fluid in a submerged chamber with addition of bicuculline (20 µM), NBQX (20 µM), and APV (100 µM) to block synaptic activity with bath temperature maintained at 31–32°C. Borosilicate glass pipettes with resistance of 4–5 MΩ were made using a Narishige PC10 vertical Micropipette Puller (Narishige International Inc.). Whole-cell patch-clamp recordings were performed from visually identified CA1 pyramidal neurons. The intracellular recording solution contained (in mM) 145 K-gluconate, 10 HEPES, 5 phosphocreatine, 2 MgCl₂, 0.1 EGTA, 2.5 Na-ATP, and 0.25 Na-GTP, osmolarity 280–290, pH 7.2. Whole-cell somatic recordings were performed using an Axopatch 200B amplifier (Molecular Devices), low-pass filtered at 2.2 kHz, and sampled at 20 kHz using a Digidata 1322A analog-to-digital interface and pClamp9 acquisition software (Molecular Devices). Seal formation and membrane rupture were done in voltage-clamp mode. Cells with series resistance more than 25 mΩ were not considered for analysis. AP firing was elicited by a series of 800 ms current pulses (from −20 to +200 pA, 10 pA increments). AP current and voltage thresholds, and passive membrane properties were determined as previously described (36). Instantaneous firing frequency was analyzed for evoked APs for all current steps at which at least 2 APs were elicited. Instantaneous firing frequency was calculated as averaged interspike intervals over the time of given current step.

Statistical and data analysis

Data are expressed as means ± sem, and the statistical significance of observed differences among groups was determined by 2-sample Student's t test. Data were tested for equal variance using Bartlett’s test with α = 0.95 and for normality using Shapiro-Wilk at a significance of 0.05. A value of P < 0.05 was regarded as statistically significant. Statistical analyses were performed and tabulated with Microsoft Excel and GraphPad Prism (Microsoft, Redmond, WA, USA). Images were processed in Adobe Photoshop CS6 and Adobe Illustrator CS6 by applying linear filters for illustration only (Adobe Systems, San Jose, CA, USA).

RESULTS

Identification of CK2 as a modulator of the FGF14:Nav complex through high-throughput screening

Previously, we have demonstrated the split-luciferase complementation assay (LCA) as a powerful and versatile method to obtain quantitative assessments of protein–protein interactions in live cells upon intracellular signaling regulation (9, 12, 20, 33, 34, 37). Briefly, the C- and N-terminal fragments of Photinus pyralis luciferase are fused to FGF14 (Cluc-FGF14) and a CD4 chimera attached to the Nav1.6 C tail (CD4-Nav1.6Ctail-Nluc) and doubly transfected into HEK293 cells, which undergo treatment with DMSO (0.5%) or small molecule kinase inhibitors. Addition of the d-luciferin substrate in the presence of reconstituted luciferase produces light, which is an indirect, but quantitative measure of protein binding (Fig. 1A, B). Through a high-throughput screening approach that we have previously described (9, 20), we identified a panel of 13 kinase pathways that control the FGF14:Nav1.6 assembly. TBB, a classic CK2 inhibitor (38), was identified to greatly suppress the FGF14:Nav1.6 interaction, but it was not chosen for further studies at the time. To test for crosstalk between CK2 and FGF14, we re-evaluated these hits and cognate compounds through the addition of a LCA dose–response panel. In these new sets of experiments, the application of 50 µM TBB resulted in nearly complete suppression of FGF14:Nav complementation, which suggests a CK2-dependent effect on FGF14 and/or the Nav1.6 C tail (Fig. 1C); however, TBB had also an effect on the full-length luciferase enzyme directly (Fig. 1D), albeit much reduced. To further validate these results, we conducted dose–response studies of CK2 inhibitor IV, an ATP-competitive inhibitor of CK2 (IC₅₀ = 9 nM) (39), which exhibited similar properties to TBB with respect to potency and cell viability but lacked effects on the reporter. We found that CK2 inhibitor IV was equally potent as compared with TBB, yet did not impair full-length luciferase enzyme activity or reduce cell viability as measured by MTT (Fig. 1E, F). In addition, cotreatment of cells with a combination of GSK-3 inhibitor XIII and CK2 inhibitor IV did not further suppress luminescence, indicating that CK2 inhibition potentially occludes the effect of GSK-3 inhibition (data not shown). These studies indicate the existence of rapid CK2-dependent regulation of the Nav channel C tail through its regulatory protein, FGF14. This rapid alteration of FGF14:Nav1.6-C-tail assembly suggests a potential mechanism of CK2-mediated regulation dependent on phosphorylation of target proteins.

Figure 1.

Inhibition of CK2 reduces FGF14:Nav1.6-Ctail binding in the split LCA. A) Schematic of the split-LCA. Left, FGF14 interacts with the C tail of Nav1.6, reconstituting intact luciferase enzyme that produces a quantitative amount of luminescence when supplied with luciferin substrate. With a kinase inhibitor X, the interaction of FGF14 with Nav1.6 is increased or decreased, increasing or reducing the amount of luminescence produced, respectively (arrows). B) Schematic of the Photinus luciferase assay. Full-length Photinus luciferase is incubated in the presence of kinase inhibitor X and the effect of the inhibitor on luminescence output is observed. C) High-through screening of over 400 kinase inhibitors at a dose of 10 µM using LCA identifies a panel of kinases as shown, including CK2 (highlighted), which significantly enhance (red) or reduce (green) FGF14:Nav1.6 complementation. Values reflect average readout of n = 2 technical replicates each. D) The Photinus luciferase assay identifies inhibitors that do not affect (black) or significantly decrease (green) full-length luciferase activity. Values reflect average readout of n = 2 technical replicates each. E) The effect of treatment with CK2 inhibitor IV on LCA complementation (black line) and Photinus luciferase signal (orange) (n = 8 for all groups). F) The effect of treatment with CK2 inhibitor IV on cell viability at 5, 25, or 50 µM assayed by MTT, compared with DMSO (n = 4 for all groups). Data are means ± sem. *P < 0.05, **P < 0.01, ***P < 0.001.

Inhibition of CK2 reduces formation of the FGF14:Nav complex

To confirm the LCA results and add an orthogonal assay, we performed coimmunoprecipitation of 6×myc-FGF14 in stably transfected HEK-Nav1.6 and Nav1.2 cells using anti-PanNav and anti-myc antibody (Fig. 2). Cells were treated for 1 h with either DMSO or TBB and the level of coimmunoprecipitation was assessed by Western blot analysis. Densitometry analysis demonstrates that treatment with 50 µM TBB significantly reduces Nav1.6 (Fig. 2A–C) (28 ± 9.0%, P = 0.02, n = 4) and Nav1.2 (Fig. 2D–F) (39 ± 5.7%, P < 0.0001, n = 7) coimmunoprecipitation with myc-FGF14, suggesting that CK2 inhibition reduces FGF14:Nav complex formation. These experimental outcomes extend the physiologic effect of CK2 inhibition on the FGF14:Nav channel complex to both Nav1.2 and Nav1.6, the 2 primary Nav channels expressed in principal neurons in the brain, extending the LCA studies, which were performed on the channel C tail alone.

Figure 2.

CK2 inhibition reduces the formation of FGF14:Nav1.6 and FGF14:Nav1.2 complexes. A, B) Cell lysates from HEK-Nav1.6 cells transiently transfected with 6×myc-FGF14 (Fgf14-1b-6×myc) and pretreated with 0.5% DMSO (control) or 50 µM TBB (A) were immunoprecipitated with an anti-myc antibody (B). C) Densitometry analysis shows that TBB reduces the FGF14:Nav1.6 complex formation compared with control (28 ± 9.0%, p=0.02, n=4 experiments). D, E) Cell lysates from HEK-Nav1.2 cells transiently transfected with 6×myc-FGF14 and pretreated with 0.5% DMSO (control) or 50 µM TBB (D) were immunoprecipitated with an anti-myc antibody (E). F) Densitometry analysis shows that TBB reduces the FGF14:Nav1.2 complex formation compared with control (39 ± 5.7%, P < 0.0001, n = 7). Data are means ± sem. *P < 0.05, ***P < 0.001.

CK2 phosphorylates FGF14 at S228 and S230 in vitro

The sequence of FGF14 contains a serine/threonine-enriched C-tail portion, which we suspected to be important for potential phosphorylation-dependent regulation of FGF14 function. Previous structure function studies supported a key role of the C tail of iFGF in modulating binding to Nav channels providing further rationale for pursuing FGF14 as the primary CK2 target (40). Using the Human Protein Reference Database, we identified 4 potential phosphorylation sites on the C tail of FGF14 that match consensus CK2 phosphorylation motifs. As CK2 is able to phosphorylate FGF1 and FGF2, we suspected that 1 or more of these consensus sites in FGF14 might undergo CK2-dependent phosphorylation. To test this hypothesis, we synthesized an FGF14 peptide fragment KPGVTPSKSTSASAIMNGGK corresponding to amino acid residues 220–239 and performed in vitro phosphorylation of the peptide with recombinant CK2 enzyme. After cleanup and desalting, the peptide was lyophilized and resolubilized in 0.1% formic acid/5% acetonitrile and analyzed by mass spectrometry with a hybrid LTQ-Orbitrap Elite (see Materials and Methods). Acquired MS2 spectra unambiguously demonstrates phosphorylation of the peptidic fragment STSASAIMNGGK at S228 (Fig. 3A) and S230 (Fig. 3B), confirming the C-tail of FGF14 as a CK2 phosphorylation site.

Figure 3.

In vitro phosphorylation followed by LC-MS/MS demonstrates that CK2 enzyme phosphorylates the FGF14 peptide KPGVTPSKSTSASAIMNGGK at S228 and S230. A) MS/MS spectrum showing fragmentation of sTSASAIMNGGK, m/z = 610.2543, z = 2, ppm = −0.9, retention time = 11.44; phosphorylation at S228. B) MS/MS spectrum showing fragmentation of STsASAIMNGGK, m/z = 610.2546, z = 2, ppm = −0.4, retention time = 11.41; phosphorylation at S230.

The FGF14^S228A/S230A mutation modifies FGF14-dependent modulation of HEK-Nav1.6 currents

Previously, we have shown that FGF14 expression in heterologous cells reduces peak Na⁺ currents; that GSK-3 inhibitors bidirectionally modulate Na⁺ currents further in the presence of FGF14, reducing current amplitudes in HEK-Nav1.6 and rescuing currents in HEK-Nav1.2; and that these changes may be due to channel cell surface availability (5, 12). To investigate possible phosphodependent regulation of FGF14 by CK2, we constructed an FGF14^S228A/S230A-GFP mutant at both of the identified phosphorylation sites of CK2 through site-directed mutagenesis of FGF14-GFP. Performing patch-clamp electrophysiology on HEK-Nav1.6 cells transfected with these plasmids, we find that FGF14^S228A/S230A-GFP broadly decreased Na⁺ current densities across a broad spectrum of voltage steps as shown in current-voltage relationships compared with FGF14-GFP (Fig. 4A–C). At −10 mV the peak Na⁺ current densities were 19.93 ± 2.76 pA/pF in FGF14-GFP (n = 19) vs. −7.6 ± 2 pA/pF (n = 10) in FGF14^S228A/S230A-GFP (P < 0.01 with Student’s t test; Fig. 4D and Table 1). Voltage dependence of activation and steady-state inactivation were not affected by the FGF14^S228A/S230A mutant (Fig. 4E, F, Table 1).

Figure 4.

Comparison of peak current densities of HEK293 Nav1.6 cells transfected with FGF14-GFP or FGF14S228A/S230A-GFP. A, B) Representative traces of voltage-gated Na⁺ currents recorded from HEK293-Nav1.6 cells transiently expressing either FGF14-GFP (A) or FGF14S228A/S230A-GFP (B). C) Current-voltage relationship of peak Na⁺ current densities for HEK293-Nav1.6 cells transiently expressing either FGF14-GFP (black circles) or FGF14S228A/S230A-GFP (purple squares). D) Bar graphs representing peak current densities at voltage obtained from HEK293 cells expressing Nav1.6 α-subunit and transfected with FGF14-GFP or FGF14S228A/S230A-GFP at voltage step of −10 mV at voltage step of −10 mV. E, F) Voltage dependences of Na⁺ current activation (E) and steady-state inactivation (F) fitted with the Boltzmann equation. **P < 0.01. Values are presented in Table 1.

TABLE 1.

Comparison of peak current densities at voltage step of −10 mV, activation and steady-state inactivation for HEK293 cells expressing Nav1.6 α-subunit and transfected with FGF14-GFP or FGF14^S228A/S230A-GFP

Condition	Peak density (pA/pF)	Activation V1/2 (mV)	kact (mV)	Inactivation V1/2 (mV)	kinact (mV)
FGF14-GFP	−19.93 ± 2.76 n = 19	−20.54 ± 0.95 n = 12	5.5 ± 05 n = 12	−53.87 ± 1.3 n = 12	6.36 ± 0.4 n = 12
FGF14S228A/S230A-GFP	−7.6 ± 2* n = 10	−21.73 ± 1.8 n = 8	4.43 ± 0.37 n = 7	−53.43 ± 0.99 n = 6	4.75 ± 0.98 n = 7

Data are means ± sem. P values obtained with Student's t test. *P < 0.01.

Inhibition of CK2 in DIV1–2 neurons reduces FGF14 accumulation in the soma and neurites

To understand the effects of CK2 inhibition in a native system, we treated DIV1–2 hippocampal neurons with DMSO or 30 μM TBB for 2 h, followed by immunostaining and confocal imaging. TBB significantly reduced FGF14 intensity (TBB: 3044 ± 57, n = 16, vs. 3760 ± 80 intensity units, n = 16, in control, P < 10⁻⁶) without significantly modifying Nav intensity in neurites (TBB: 3686 ± 77, n = 16, vs. 3854 ± 65 intensity units, n = 16, in control, P = 0.25), and significantly reduced FGF14 intensity in the soma (TBB: 4152 ± 133, n = 19, vs. 5943 ± 212 intensity units, n = 20, in control, P < 10⁻⁶) while modestly decreasing Nav accumulation in the soma of DIV1–2 neurons (TBB: 5547 ± 149, n = 19, vs. 6010 ± 157 intensity units, n = 20, in control, P = 0.04) (Fig. 5A–H). These effects resulted in a profound increase in the Nav/FGF14 ratio in both neurites (TBB: 1.21 ± 0.02, n = 16, vs. 1.03 ± 0.02, n = 16, in control, P < 10⁻⁶) and soma (TBB: 1.35 ± 0.04, n = 19, vs. 1.02 ± 0.02, n = 20, in control, P < 10⁶) of DIV1–2 neurons (Fig. 5I–P).

Figure 5.

CK2 inhibition reduces FGF14 accumulation in neurites and somas of DIV1–2 hippocampal neurons. ×63 zoom confocal images of DIV1–2 hippocampal neurons treated with 0.25% DMSO control (A–D) or 30 μM TBB (E–H) stained for PanNav (A, E), FGF14 (B, F), and MAP2 (C, G), along with overlay images (D, H). Region used for background intensity correction denoted as blue dotted circle. I, J) Quantification of average intensity along the length of neurites for Nav (I) and FGF14 (J) with DMSO-treated (white, n = 16) and TBB-treated (purple, n = 16) neurons (example in yellow dotted box). TBB significantly reduced FGF14 intensity compared with control (TBB: 3044 ± 57, n = 16, vs. 3760 ± 80 intensity units, n = 16, in control, P < 10⁻⁶). K) Nav/FGF14 ratio for neurites (TBB: 1.21 ± 0.02, n = 16, vs. 1.03 ± 0.02, n = 16, in control, P < 10⁻⁶). L) Length of neurites. M, N) Quantification of average intensity in a 6-pixel line segment spanning the diameter of the neuronal soma for Nav (M) and FGF14 (N) with DMSO (white, n = 20) and TBB (purple, n = 19) treated neurons (example in red dotted box). TBB significantly reduced FGF14 intensity (TBB: 4152 ± 133, n = 19, vs. 5943 ± 212 intensity units, n = 20, in control, P < 10⁻⁶) and modestly reduced PanNav intensity (TBB: 5547 ± 149, n = 19, vs. 6010 ± 157 intensity units, n = 20, in control, P = 0.04) vs. control. O) Nav/FGF14 ratio for soma (TBB: 1.35 ± 0.04, n = 19, vs. 1.02 ± 0.02, n = 20, in control, P < 10⁻⁶). P) Diameter of soma. Box plots shown with means ± 1.5 IQR, whiskers min–max. *P < 0.05, ***P < 10⁻⁶.

Inhibition of CK2 reduces transient Na⁺ currents

To investigate the functional effects of inhibiting CK2 in a native system, we first used patch-clamp electrophysiology to perform whole-cell patch-clamp recordings on DIV1-3 primary hippocampal neurons treated (2 h) with 0.5% DMSO or 30 μM TBB. Our goal was to isolate transient Na⁺ currents (I_Na+) at a developmental stage with immature neurites to permit a good space clamp with limited voltage-errors. Based on the modulatory action of Nav channels by FGF14 (4) and our confocal imaging at DIV1, we posited that CK2 inhibition would diminish Nav channel-mediated responses and alter kinetics. As expected, TBB-treated neurons exhibited a profound reduction in I_Na+ current amplitude (Table 2 and Fig. 6A–C) compared with control (TBB: −18.4 ± 3.7, n = 7, vs. −56.6 ± 3.9 pA/pF, n = 8, in control, P < 10⁻⁴). Furthermore, treatment with TBB did not affect Nav channel activation (Table 2 and Fig. 6D), but significantly shifted the voltage dependence of steady-state inactivation to hyperpolarizing values (V_1/2 in TBB: 78.6 ± 1.9, n = 6, vs. −59.2 ± 0.8 mV, n = 6, in control, P < 10⁻⁴) (Table 2 and Fig. 6E), indicating that inhibition of CK2 activity decreases availability of Nav channels. In addition, analysis of peak current density values from a voltage step of −10 mV, preceded by a prepulse of −110 mV again showed significant reduction in neuronal sodium currents in the TBB group compared with DMSO control (−65.84 ± 5.61 pA/pF in the DMSO group, n = 6 vs. −25.08 ± 5.29 pA/pF in the TBB group, P = 0.0004 with Student’s t test). Importantly, these phenotypes reproduce the effect of the FGF14^F145S dominant-negative mutation on I_Na+ (4), suggesting a potential convergence of FGF14 and CK2 on Nav channel activity. Amplitude and kinetics of I_Na+ are key molecular determinants of neuronal excitability, as they set AP threshold and regulate firing patterns. Thus, the profound changes observed in neurons treated with TBB could be predictors of impaired intrinsic excitability.

TABLE 2.

Effects of TBB on I_Na+ in primary hippocampal neurons

INa+ properties	DMSO	30 μM TBB	P value
Peak density (pA/pF)	−56.6 ± 3.9 (8)	−18.4 ± 3.7 (7)	<0.0001***
Activation V1/2 (mV)	−23.5 ± 1.7 (8)	−24.04 ± 2.6 (6)	0.86
kact (mV)	4.4 ± 0.5 (8)	5.02 ± 1.2 (6)	0.61
Inactivation V1/2 (mV)	−59.2 ± 0.8 (6)	−78.6 ± 1.9 (6)	0.0001***
kinact (mV)	5.4 ± 0.2 (6)	5.9 ± 1.1 (6)	0.66

Numbers in parentheses denote n of data. Data are means ± sem. P values obtained with Student's t test. ***P < 0.001.

Figure 6.

TBB suppresses I_Na+ current densities in primary hippocampal neurons. A) Representative traces of transient sodium currents (I_Na+) recorded in the presence of DMSO (left) and 30 μM TBB (right). B) Current-voltage relationships for neuronal I_Na+ in the presence of DMSO and 30 μM TBB. Inhibition of CK2 reduces I_Na+ peak amplitude across broad range of given voltages. C) Bar graphs representing average peak I_Na+ densities at a voltage step of −10 mV. Inhibition of CK2 with 30 μM TBB significantly reduces peak amplitude of neuronal I_Na+ compared with group treated with DMSO (TBB: −18.4 ± 3.7, n = 7, vs. −56.6 ± 3.9 pA/pF, n = 8, in control, P < 10⁻⁴). ***P < 0.001 with Student’s t test. D) Voltage-dependence of activation of neuronal Nav channels in the presence of DMSO and 30 μM TBB. E) Voltage dependence of steady-state inactivation of Nav channels in the presence of DMSO and 30 μM TBB. Inhibition of CK2 with 30 μM TBB induces a significant hyperpolarizing shift of steady-state inactivation compared with DMSO control group (V_1/2 in TBB: −78.6 ± 1.9 mV, n = 6, vs. −59.2 ± 0.8 mV, n = 6, in control, P < 10⁻⁴).

Inhibition of CK2 induces time-dependent redistribution of FGF14 polarity in mature neurons

To expand our study of CK2 inhibition to mature neurons, we treated primary rat hippocampal neurons (DIV21) with DMSO or 50 µM TBB for 1, 2, or 8 h, followed by immunostaining and confocal imaging. We show that TBB treatment significantly reduced the expression of FGF14 at the AIS within 1 h of treatment [TBB: 1.47 ± 0.13, n = 14, vs. 2.10 ± 0.18 intensity units (×10⁵), n = 22, in control, P = 0.01] (Fig. 7A–C) and induces a pronounced redistribution of FGF14 immunostaining from the AIS to the dendrites reaching the peak at 8 h (TBB: 0.97 ± 0.09, n = 7, ratio of axonal FGF14 to dendritic FGF14 intensity vs. 4.09 ± 0.27, n = 22, in control, P < 0.0001) (Fig. 7D). Within this short time frame, treatment with TBB did not appear to affect the axodendritic distribution or fluorescence intensity of Nav (Fig. 7F, G) and/or β-IV spectrin (Fig. 7D), an axonal protein closely associated with the Nav complex. These data suggest a model where changes in CK2 activity might convey a rapid intracellular signal that alters the subcellular distribution of FGF14 within the axodendritic compartment (Fig. 7E). Consequently, alterations in FGF14 distribution can be expected to impact Na⁺ channel activity and neuronal firing.

Figure 7.

CK2 inhibition induces a time-dependent loss of FGF14 axonal polarity. A) Confocal images of DIV21 primary hippocampal neurons immunolabeled for MAP2 (i–iv), FGF14 (v–viii), and βIV-Spectrin (ix–xii), along with overlay images (xiii–xvi). Region used for background intensity correction denoted as blue dotted circle. Upon increasing exposure to 50 µM TBB, FGF14 steadily decreases from the AIS and accumulates in the dendrites. B, C) Quantification of FGF14 redistribution in AIS (B) and dendrites (C) upon increasing exposure to TBB. Axons: TBB (1 h): 1.47 ± 0.13 intensity units (×10⁵), n = 14 neurites, n = 1 experiment; TBB (2 h): 1.27 ± 0.14, n = 8, n = 1; TBB (8 h): 1.34 ± 0.07, n = 7, n = 3; control: 2.10 ± 0.18, n = 22, N = 4; P = 0.01, 0.01, 0.026 for 1, 2, and 8 h, respectively. Dendrites: TBB (1 h): 0.49 ± 0.04 intensity units (×10⁵), n = 14; TBB (2 h): 0.45 ± 0.03, n = 8, n = 1; TBB (8 h): 1.45 ± 0.14, n = 7, n = 3; control: 0.53 ± 0.04, n = 22, n = 4; P = 0.56, 0.31, < 0.0001 for 1, 2, and 8 h, respectively. D) The axonal polarity of FGF14 (pixel intensity in AIS/ pixel intensity in dendrites) reverses with increasing time exposure to TBB (8 h): TBB: 0.97 ± 0.09, n = 7, n = 3, ratio of FGF14 to PanNav intensity vs. 4.09 ± 0.27, n = 22, n = 4 in control, P < 0.0001). Data are means ± sem (n = 8–13 per condition). *P < 0.05; **P < 0.01; ***P < 0.001. E) Model showing rapid redistribution of FGF14 from the AIS compartment to the dendritic compartment upon inhibition of CK2 by TBB. F) Confocal images of DIV21 primary hippocampal neurons immunolabeled for PanNav (i, ii), βIV-Spectrin (iii, iv), and MAP2 (v, vi). G) The distribution of PanNav in the axon is unchanged upon treatment with 8 h TBB compared with DMSO. Data are means ± sem (n = 8 per condition, n = 2 experiments).

Inhibition of CK2 impairs intrinsic excitability

To examine intrinsic firing properties of neurons in response to modulation of CK2 activity, we performed whole-cell patch-clamp recordings in current-clamp mode from visually identified CA1 hippocampal pyramidal neurons in acute brain slices. To rule out network activity, recordings were performed in the presence of the GABAergic and glutamatergic synaptic blockers bicuculline (20 µM), NBQX (20 µM), and APV (100 µM). Incubation with TBB (40–60 min; 30 µM) significantly decreases the maximum number of APs (Fig. 8A–C) and lowered instantaneous firing frequency across the entire spectrum of injected current (Fig. 8D). Furthermore, the reduction in AP number was associated with an increase in the AP current threshold (TBB: 87.1 ± 15.6, n = 7, vs. 36 ± 6.5 pA, n = 10, in control, P = 0.001), and no changes in membrane capacitance, membrane resistance, and AP voltage threshold were detected ruling out a major role of TBB on other conductances (i.e., K⁺ channels) and on membrane passive properties (Table 2). Though indirectly, these results indicate that TBB might impair neuronal excitability by suppressing I_Na+ and decreasing Nav channel availability, but do not indicate whether FGF14 had any role in this regulation. We posited that if FGF14 mediated TBB regulation of neuronal firing, then Fgf14^−/− mice would mimic the effect of TBB treatment on neuronal intrinsic excitability and prevent any further effect of the inhibitor. We tested this hypothesis by measuring intrinsic firing properties from CA1 pyramidal neurons in Fgf14^−/− mice in the presence and absence of TBB.

Figure 8.

CK2 inhibition suppresses intrinsic excitability of CA1 pyramidal neurons. A, B) Superimposed representative current-clamp traces from −20, 0, 50, 100, and 150 pA injected current steps for CA1 hippocampal pyramidal neurons after 30–40 min of slice preincubation with either DMSO (A) or TBB 30 µM (B). C, D) Input–output curves of either number of APs (C) or averaged instantaneous firing frequency (D) vs. injected current stimuli recorded in hippocampal slices after preincubation with DMSO or TBB 30 µM. P values obtained with Student's t test. *P < 0.05; **P < 0.01; ***P < 0.001.

Rapid inhibition of CK2 impairs intrinsic excitability in an FGF14-dependent manner

As shown in Fig. 9, firing properties of CA1 pyramidal neurons in Fgf14^−/− were greatly disrupted compared with Fgf14^+/+ control littermates (Fig. 8). This is a phenotype that we expected considering that in hippocampal neurons the FGF14^F145S dominant-negative mutation disrupts intrinsic firing (along with suppressing I_Na+). Maximum firing and instantaneous frequency were suppressed in CA1 pyramidal neurons from Fgf14^−/− mice and current thresholds were similarly increased (67.5 ± 10.6 pA, n = 8, in Fgf14^−/− vs. 36 ± 6.5 pA, n = 10, in WT, P = 0.018) (Fig. 9 and Table 3). Notably, incubation with 30 µM TBB did not further aggravate the phenotypes compared with vehicle (Fig. 9 and Table 3) supporting the hypothesis that Fgf14^−/− mice mimic a CK2-dependent phenotype. Such an experimental outcome confirms convergence of CK2 on FGF14 and establishes a novel mechanism of regulation of neuronal firing through modulation of protein–protein interactions at the Nav channel complex.

Figure 9.

Genetic deletion of Fgf14 occludes TBB-dependent regulation of intrinsic excitability of CA1 pyramidal neurons. A, B) Superimposed representative current-clamp traces from −20, 0, 50, 100, and 150 pA injected current steps for Fgf14^−/− CA1 hippocampal pyramidal neurons after 30–40 min of slice preincubation with either DMSO (A) or TBB 30 µM (B). C, D) Input–output curves of either number of APs (C) or averaged instantaneous firing frequency (D) vs. injected current steps recorded in Fgf14^−/− hippocampal slices after preincubation with DMSO or TBB 30 µM. All the results for C and D were statistically insignificant through all the given stimuli. P > 0.05 with Student’s t test.

TABLE 3.

Effects of TBB and Fgf14^−/− on active/passive properties and AP thresholds in CA1 pyramidal neurons

Neuronal active and passive properties	DMSO, WT (i) n = 10 neurons	TBB, WT (ii) n = 7 neurons	DMSO, Fgf14−/− (iii) n = 8 neurons	TBB, Fgf14−/− (iv) n = 6 neurons
Membrane capacitance (pF)	84.7 ± 7.7	88.61 ± 16.1 P = 0.80 vs. i	104.5 ± 9.6 P = 0.12 vs. i	95.0 ± 11.9 P = 0.76 vs. ii P = 0.51 vs. iii
Membrane input resistance (MΩ)	223.9 ± 22.6	174.9 ± 27.7 P = 0.14 vs. i	165.5 ± 16.9 P = 0.07 vs. i	178.8 ± 18.9 P = 0.91 vs. ii P = 0.57 vs. iii
Time constant (ms)	17.8 ± 1.3	14.1 ± 1.9 P = 0.10 vs. i	17.0 ± 1.2 P = 0.66 vs. i	16.2 ± 1.3 P = 0.40 vs. ii P = 0.90 vs. iii
Current threshold (pA)	36 ± 6.5	87.1 ± 15.6 P = 0.001 vs. i**	67.5 ± 10.6 P = 0.02 vs. i*	66.6 ± 10.5 P = 0.32 vs. ii P = 0.95 vs. iii
Voltage threshold (mV)	−51.2 ± 2.0	−45.4 ± 3.6 P = 0.12 vs. i	−50.5 ± 2.1 P = 0.81 vs. i	−49.2 ± 2.2 P = 0.41 vs. ii P = 0.58 vs. iii

Data are means ± sem. P values obtained with Student's t test. *P < 0.05, **P < 0.01.

DISCUSSION

Our study provides evidence for a novel mechanism of regulation of the Nav channel complex and excitability through a CK2-FGF14 axis. With a combination of luminescence-based screening, in vitro phosphorylation assays, high-resolution mass spectrometry, coimmunoprecipitation, and confocal imaging we demonstrate that CK2 phosphorylates FGF14 and that inhibition of CK2 greatly suppresses binding of FGF14 to Nav channels in a heterologous system. Patch-clamp electrophysiology in dissociated neurons elicited a strong suppression of Na⁺ currents due to TBB, and recordings in brain slices further corroborates the hypothesis through a TBB-dependent suppression of intrinsic excitability, mimicked and occluded by Fgf14 deletion. All combined these results support the existence of a dynamic and fast-responding CK2-FGF14 axis that adds to the existing complexity of the CK2 signaling pathway in neurons (29, 31).

Short-term changes in CK2 activity, as described in this study can influence Na⁺ currents and consequently neuronal excitability, presumably through decreased phosphorylation and dissociation of FGF14 from the Nav channel complex (Fig. 7). As opposed to direct phosphorylation of Nav channels, phosphorylation of FGF14 by CK2 might represent a rapid mode for neurons to adapt their outputs in response to cellular stimuli without directly intervening on the Nav channel. Given CK2 clustering at the AIS and its tight association with Ankyrin-G (25, 26), we expect posttranslational modifications through CK2 phosphorylation to have important consequences for Nav complex function. CK2 inhibition affects distribution of FGF14 in the soma, dendrites and axons in a developmentally regulated manner; for instance, at early developmental stages (DIV1), the somatic pool is more affected by CK2 inhibition compared with neurons with established polarity. Interestingly, the effect of CK2 inhibition in disrupting FGF14 polarity is more pronounced at an earlier time for mature axons as compared with mature dendrites (Fig. 7B, C); this delay may be due to temporary somatic sequestration of FGF14 and/or slower diffusion of FGF14 from the somatic to the dendritic compartments. This study is, to our knowledge, the first demonstration that the interaction of CK2 with iFGFs can affect the functional properties of neurons. At the AIS, temporal segregation and target specificity in CK2 signaling might have implications for establishing polarity and directional signaling in developing and mature networks through activity-dependent synaptic plasticity (rapid response via FGF14 phosphorylation) or homeostatic remodeling (via direct Nav channel phosphorylation) (18, 41–43). Additionally, modulation of axodendritic polarity of FGF14 by CK2 activity (Figs. 5 and 7) may have important consequences for the function of axon-carrying dendrites, which serve as a “privileged” channel for reception of excitatory synaptic input (44).

Together, our findings show that CK2 phosphorylates FGF14 in vitro at S228 and S230, and inhibition of CK2 via TBB reduces the interaction of FGF14 with Nav1.6-Ctail in LCA and also inhibits the association between 6× myc-FGF14 and Nav1.6/Nav1.2. Furthermore, CK2 inhibition reduces FGF14 expression, reduces the axodendritic polarity of FGF14, and disrupts Nav-mediated currents by reducing peak current densities and shifting the voltage dependence of steady-state inactivation curve leftward, which suggests a combined effect of CK2 on regulating the membrane trafficking pool as well as the biophysical properties of the channel. Both changes are consistent with a reduced interaction of FGF14 with Nav channels (Figs. 5 and 7) and are remarkably similar to the phenotypes observed in hippocampal neurons expressing the FGF14^F145S dominant-negative mutation (4). Furthermore, CK2 inhibition decreases the total number of evoked AP and instantaneous AP frequency while increasing AP current threshold, a phenotype that can be reconciled with a decrease in Nav channel number and/or availability; though other channels that could contribute to AP current threshold could also be affected (45). The increase in current threshold upon CK2 inhibition (Table 2) without concomitant changes in membrane capacitance, resistance, and τ indicate that the effect of CK2 is likely directed against Nav channels, and not on other channel types. Furthermore, Fgf14^−/− recapitulates the selective increase in current threshold seen in TBB-treated wild-type mice, and that in Fgf14^−/− mice TBB has no further effects on any AP and firing-related phenotypes (Table 3), indicating a clear occlusion and potential convergence of the 2 mechanisms. The absence of further suppression of firing in Fgf14^−/− mice treated with TBB (relative to treatment with control) suggests that CK2 activity through pharmacological inhibition by TBB is mediated by the presence of FGF14.

As previously shown, pharmacological inhibition of GSK-3 reduced the assembly of the FGF14:Nav complex, modifies Na⁺ currents, and alters the distribution and localization of FGF14/Nav in hippocampal neurons (12). Interestingly, the C-tail of FGF14 also contains a putative GSK-3 phosphorylation site, S226, of the S/TXXXS/T motif, and it is known that GSK-3 sites are often accompanied by priming phosphorylation events of flanking sites by priming kinases (23, 46). Additionally, CK2 is known to participate in multisite phosphorylation events with partner kinases across the entire genome, particularly with proline-directed kinases such as GSK-3 (47). This suggests that S228 and S230, which are shown in this study to be phosphorylated by CK2 and control the modulation of Na⁺ currents by FGF14, could be priming sites for GSK-3 phosphorylation events. Therefore, phosphorylation of FGF14 by CK2 could prime the Nav channel complex for GSK-3 conveying temporal and spatial specifications for Nav activity and serving as an amplification loop and coincident detector for the GSK-3 pathway. Notably, FGF14^S228A/S230A-GFP further suppresses Nav1.6-mediated currents (Fig. 4 and Table 1) with a mechanism that resembles the activity of GSK-3 inhibitors on FGF14 (12), further supporting convergence of the 2 pathways on the FGF14:Nav complex. Though modulation of Na⁺ currents differs between native and heterologous systems, heterologous systems are invaluable models to study iFGF and specific Nav isoforms in isolation and generate mechanistic hypotheses to explain complex behaviors in the native system. Thus, crosstalk between these 2 pathways on the Nav channel complex deserves further investigation. These new data indicate that the FGF14 sites phosphorylated by CK2 contribute to the regulation of Nav1.6 and are likely part of the complex regulation of excitability by CK2 in the native system.

Previously, disruption of FGF14 has been observed in genomewide association studies to be a factor in major depressive disorder (7), and the FGF14(F145S) mutation, implicated in hereditary spinocerebellar ataxia, has been characterized as a dominant-negative mutation that mimics the ataxic and dystonic phenotype seen in fgf14^−/− mice (4, 16, 48, 49). Likewise, both excesses and deficits of CK2 have been linked to neurologic and psychiatric disorders, with hyperphosphorylation of α-synuclein in Lewy body disease (50) and deficiencies in syntaxin 1 phosphorylation in the schizophrenic prefrontal cortex (51). Overall, collective evidence suggests that FGF14 might be a kinase scaffold and molecular shuttle that neurons use as a route to convey intracellular signaling to Nav channel complexes for fine-tuning regulation of neuronal excitability.

Glossary

AIS

axonal initial segment

AP

action potential

APV

(2R)-amino-5-phosphonovaleric acid

CA1

cornu ammonis area 1

CD4

cluster of differentiation 4

CK2

casein kinase 2

DIV

days in vitro

FGF14

fibroblast growth factor 14

GSK-3

glycogen synthase kinase 3

HEPES

4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HEK293

human embryonic kidney 293

iFGF

intracellular fibroblast growth factor

LCA

luciferase complementation assay

Nav

voltage-gated sodium

NBQX

2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo[f]quinoxaline-2,3-dione

NIH

U.S. National Institutes of Health

TBB

4,5,6,7-tetrabromobenzotriazole

TCEP

tris-(2-carboxyethyl)phosphine

UTMB

University of Texas Medical Branch